The interaction of odorants with olfactory receptors on the apical cilia of olfactory neurons is the first step in our perception of smell. The large number (e.g., approximately 400 in humans) and structural diversity of the opsin-like GPCRs that function as olfactory receptors underlies our ability to detect and discriminate a vast number of volatile compounds (Buck and Axel, 1991, Fuchs et al., 2001). Although there appear to be several hundred different olfactory neuron cell types in humans—each expressing a different olfactory receptor gene (Chess et al., 1994; Malnic et al., 1999; Touhara et al., 1999)—several lines of evidence argue that signal transduction in olfactory neurons involves a common downstream signaling pathway that involves the second messenger cAMP. Many odorants have been shown to elicit an increase in cAMP in olfactory neurons (Breer, 1993). Moreover, mouse knockout studies with the Gs-like G protein Golf (Jones and Reed, 1989; Belluscio et al., 1998), type III adenylyl cyclase (Bakalyar and Reed, 1990; Wong et al., 2000), and an olfactory CNG channel subunit (Dhallan et al., 1990; Brunet et al., 1996) have demonstrated that these signaling molecules play an obligate role in olfactory signal transduction. In summary, olfactory signal transduction is believed to involve the following steps: (1) an olfactory stimulus activates an olfactory receptor; (2) the activated olfactory receptor activates Golf; (3) type III adenylyl cyclase is activated by Golf and catalyzes cAMP synthesis; (4) the olfactory CNG channel is activated by cAMP; (4) ion flux through the activated olfactory CNG channel depolarizes the olfactory neuron and initiates the generation of action potentials (Gold, 1999).
CNG channels are composed of structurally related subunits with six transmembrane segments and C-terminal nucleotide-binding domains (Zagotta and Siegelbaum, 1996). The first subunit, OCNC1, of the olfactory neuron-specific CNG channel to be identified was cloned from rat olfactory cDNA based on its homology to retinal CNG channel subunits (Dhallan et al., 1990). Discrepancies between the native olfactory CNG channel conductance and sensitivity to cAMP and cGMP and the properties of heterologously expressed OCNC1 channels motivated the search for additional CNG channel subunits. Two additional subunits, OCNC2 and β1b (a splice variant of a rod photoreceptor CNG channel subunit), were subsequently identified by homology-based cloning from rodent olfactory cDNA (Liman and Buck, 1994; Bradley et al., 1994; Sautter et al., 1998; Bonigk et al., 1999). Histochemical experiments demonstrated that these molecules are selectively expressed in the apical cilia of olfactory neurons. And heterologous coexpression of the olfactory CNG subunits produced channels with native-like properties (Sautter et al., 1998; Bonigk et al., 1999). Therefore, the olfactory CNG channel may be composed of three different subunits. (OCNC1 is also called CNG2, CNGC4, CNCα3, and olfactory CNG channel α; OCNC2 is also called CNG5, CNCα4, and olfactory CNG channel β; β1b is also called CNCβ1b and CNG4.3.)
Olfactory CNG channels are compelling targets for smell modulators. First, the channels are localized to the apical cilia of olfactory neurons and exposed on the external surface of the olfactory epithelium. Therefore, channel activators do not have to cross cell membranes. Second, olfactory CNG channels can be functionally expressed in heterologous cells (Sautter et al., 1998; Bonigk et al., 1999; Qu et al., 2000). Third, calcium-imaging-based assays, which are amenable to high-throughput screening with automated fluorometric instrumentation, can be used to measure CNG channel activity because these channels are permeable to Ca2+. Indeed, calcium imaging has been used to record odorant-dependent activation of olfactory neurons in primary culture (Hirono et al., 1992; Ma and Shepherd, 2000), and a calcium-dye-based assay was recently developed for rat OCNC1 expressed in heterologous cells (Rich et al., 2001). Fourth, these channels are gated by cyclic nucleotides—i.e., they are allosterically activated by small molecules. Finally, several small molecules have been identified that may modulate smell by directly inhibiting CNG channel activity (Kurahashi et al., 1994). And there is ample precedent for small-molecules that activate non-ionotropic receptor ion channels such as dihydropyridine agonists of L-type Ca2+ channels (Hockerman et al., 1997).
Rodent olfactory CNG channels have been expressed in Xenopus oocytes and HEK-293 cells, and they have been functionally characterized by direct measurement of the electrical properties of CNG channel-containing membranes (Sautter et al., 1998; Bonigk et al., 1999; Qu et al., 2000). However, electrophysiological assays are not amenable to high-throughput screening. To overcome this limitation, the present invention describes a cell-based fluorometric assay that is amenable to automated high-throughput screening for olfactory CNG channel modulators. This assay uses fluorescent calcium or membrane potential dyes to quantitate ion flux through olfactory CNG channels expressed in HEK-293 cells following elevation of cAMP levels caused by the adenylyl cyclase activator forskolin. In addition, the invention identifies nucleic acid sequences that encode human olfactory CNG channel subunits, which had not been identified prior to this invention.